dendropy.calculate.treecompare: Distances and Comparison Between Trees

Statistics, metrics, measurements, and values calculated between two trees.

dendropy.calculate.treecompare.euclidean_distance(tree1, tree2, edge_weight_attr='length', value_type=<class 'float'>, is_bipartitions_updated=False)[source]

Returns the Euclidean distance (a.k.a. Felsenstein’s 2004 “branch length distance”) between two trees based on edge_weight_attr.

Trees need to share the same TaxonNamespace reference. The bipartition bitmasks of the trees must be correct for the current tree structures (by calling Tree.encode_bipartitions method) or the is_bipartitions_updated argument must be False to force recalculation of bipartitions.

Parameters:

  • tree1 (Tree object) – The first tree of the two trees being compared. This must share the same TaxonNamespace reference as tree2 and must have bipartitions encoded.

  • tree2 (Tree object) – The second tree of the two trees being compared. This must share the same TaxonNamespace reference as tree1 and must have bipartitions encoded.

  • edge_weight_attr (string) – Name of attribute on edges of trees to be used as the weight.

  • is_bipartitions_updated (bool) – If False (default), then the bipartitions on both trees will be updated before comparison. If True then the bipartitions will only be calculated for a Tree object if they have not been calculated before, either explicitly or implicitly.

Returns:

d (int) – The Euclidean distance between tree1 and tree2.

Notes

Encoding bipartitions may collapse an unrooted basal bifurcation or suppress unifurcations, modifying the tree. Alternatively, call Tree.encode_bipartitions with desired options then pass is_bipartitions_updated=True.

Examples

import dendropy
from dendropy.calculate import treecompare
tns = dendropy.TaxonNamespace()
tree1 = tree.get_from_path(
        "t1.nex",
        "nexus",
        taxon_namespace=tns)
tree2 = tree.get_from_path(
        "t2.nex",
        "nexus",
        taxon_namespace=tns)
tree1.encode_bipartitions()
tree2.encode_bipartitions()
print(treecompare.euclidean_distance(tree1, tree2))
dendropy.calculate.treecompare.false_positives_and_negatives(reference_tree, comparison_tree, is_bipartitions_updated=False)[source]

Counts and returns number of false positive bipar (bipartitions found in comparison_tree but not in reference_tree) and false negative bipartitions (bipartitions found in reference_tree but not in comparison_tree).

Trees need to share the same TaxonNamespace reference. The bipartition bitmasks of the trees must be correct for the current tree structures (by calling Tree.encode_bipartitions method) or the is_bipartitions_updated argument must be False to force recalculation of bipartitions.

Parameters:

  • reference_tree (Tree object) – The first tree of the two trees being compared. This must share the same TaxonNamespace reference as tree2 and must have bipartitions encoded.

  • comparison_tree (Tree object) – The second tree of the two trees being compared. This must share the same TaxonNamespace reference as tree1 and must have bipartitions encoded.

  • is_bipartitions_updated (bool) – If False (default), then the bipartitions on both trees will be updated before comparison. If True then the bipartitions will only be calculated for a Tree object if they have not been calculated before, either explicitly or implicitly.

Returns:

t (tuple(int)) – A pair of integers, with first integer being the number of false positives and the second being the number of false negatives.

Notes

Encoding bipartitions may collapse an unrooted basal bifurcation or suppress unifurcations, modifying the tree. Alternatively, call Tree.encode_bipartitions with desired options then pass is_bipartitions_updated=True.

Examples

import dendropy
from dendropy.calculate import treecompare
tns = dendropy.TaxonNamespace()
tree1 = tree.get_from_path(
        "t1.nex",
        "nexus",
        taxon_namespace=tns)
tree2 = tree.get_from_path(
        "t2.nex",
        "nexus",
        taxon_namespace=tns)
tree1.encode_bipartitions()
tree2.encode_bipartitions()
print(treecompare.false_positives_and_negatives(tree1, tree2))
dendropy.calculate.treecompare.find_missing_bipartitions(reference_tree, comparison_tree, is_bipartitions_updated=False)[source]

Returns a list of bipartitions that are in reference_tree, but not in comparison_tree.

Trees need to share the same TaxonNamespace reference. The bipartition bitmasks of the trees must be correct for the current tree structures (by calling Tree.encode_bipartitions method) or the is_bipartitions_updated argument must be False to force recalculation of bipartitions.

Parameters:

  • reference_tree (Tree object) – The first tree of the two trees being compared. This must share the same TaxonNamespace reference as tree2 and must have bipartitions encoded.

  • comparison_tree (Tree object) – The second tree of the two trees being compared. This must share the same TaxonNamespace reference as tree1 and must have bipartitions encoded.

  • is_bipartitions_updated (bool) – If False (default), then the bipartitions on both trees will be updated before comparison. If True then the bipartitions will only be calculated for a Tree object if they have not been calculated before, either explicitly or implicitly.

Returns:

s (list[|Bipartition|]) – A list of bipartitions that are in the first tree but not in the second.

Notes

Encoding bipartitions may collapse an unrooted basal bifurcation or suppress unifurcations, modifying the tree. Alternatively, call Tree.encode_bipartitions with desired options then pass is_bipartitions_updated=True.

dendropy.calculate.treecompare.mason_gamer_kellogg_score(tree1, tree2, is_bipartitions_updated=False)[source]

Mason-Gamer and Kellogg. Testing for phylogenetic conflict among molecular data sets in the tribe Triticeae (Gramineae). Systematic Biology (1996) vol. 45 (4) pp. 524

dendropy.calculate.treecompare.robinson_foulds_distance(tree1, tree2, edge_weight_attr='length')[source]

DEPRECATED: Use :func:symmetric_difference for the common unweighged Robinson-Fould’s distance metric (i.e., the symmetric difference between two trees) :func:weighted_robinson_foulds_distance or for the RF distance as defined by Felsenstein, 2004.

dendropy.calculate.treecompare.symmetric_difference(tree1, tree2, is_bipartitions_updated=False)[source]

Returns unweighted Robinson-Foulds distance between two trees.

Trees need to share the same TaxonNamespace reference. The bipartition bitmasks of the trees must be correct for the current tree structures (by calling Tree.encode_bipartitions method) or the is_bipartitions_updated argument must be False to force recalculation of bipartitions.

Parameters:

  • tree1 (Tree object) – The first tree of the two trees being compared. This must share the same TaxonNamespace reference as tree2 and must have bipartitions encoded.

  • tree2 (Tree object) – The second tree of the two trees being compared. This must share the same TaxonNamespace reference as tree1 and must have bipartitions encoded.

  • is_bipartitions_updated (bool) – If False, then the bipartitions on both trees will be updated before comparison. If True then the bipartitions will only be calculated for a Tree object if they have not been calculated before, either explicitly or implicitly.

Returns:

d (int) – The symmetric difference (a.k.a. the unweighted Robinson-Foulds distance) between tree1 and tree2.

Notes

Encoding bipartitions may collapse an unrooted basal bifurcation or suppress unifurcations, modifying the tree. Alternatively, call Tree.encode_bipartitions with desired options then pass is_bipartitions_updated=True.

Examples

from dendropy import TaxonNamespace, Tree
from dendropy.calculate import treecompare
tns = TaxonNamespace()
tree1 = Tree.get_from_path(
        "t1.nex",
        "nexus",
        taxon_namespace=tns)
tree2 = Tree.get_from_path(
        "t2.nex",
        "nexus",
        taxon_namespace=tns)
tree1.encode_bipartitions()
tree2.encode_bipartitions()
print(treecompare.symmetric_difference(tree1, tree2))
dendropy.calculate.treecompare.unweighted_robinson_foulds_distance(tree1, tree2, is_bipartitions_updated=False)[source]

Alias for symmetric_difference().

Notes

Encoding bipartitions may collapse an unrooted basal bifurcation or suppress unifurcations, modifying the tree. Alternatively, call Tree.encode_bipartitions with desired options then pass is_bipartitions_updated=True.

dendropy.calculate.treecompare.weighted_robinson_foulds_distance(tree1, tree2, edge_weight_attr='length', is_bipartitions_updated=False)[source]

Returns weighted Robinson-Foulds distance between two trees based on edge_weight_attr.

Trees need to share the same TaxonNamespace reference. The bipartition bitmasks of the trees must be correct for the current tree structures (by calling Tree.encode_bipartitions method) or the is_bipartitions_updated argument must be False to force recalculation of bipartitions.

Parameters:

  • tree1 (Tree object) – The first tree of the two trees being compared. This must share the same TaxonNamespace reference as tree2 and must have bipartitions encoded.

  • tree2 (Tree object) – The second tree of the two trees being compared. This must share the same TaxonNamespace reference as tree1 and must have bipartitions encoded.

  • edge_weight_attr (string) – Name of attribute on edges of trees to be used as the weight.

  • is_bipartitions_updated (bool) – If False (default), then the bipartitions on both trees will be updated before comparison. If True then the bipartitions will only be calculated for a Tree object if they have not been calculated before, either explicitly or implicitly.

Returns:

d (float) – The edge-weighted Robinson-Foulds distance between tree1 and tree2.

Notes

Encoding bipartitions may collapse an unrooted basal bifurcation or suppress unifurcations, modifying the tree. Alternatively, call Tree.encode_bipartitions with desired options then pass is_bipartitions_updated=True.

Examples

import dendropy
from dendropy.calculate import treecompare
tns = dendropy.TaxonNamespace()
tree1 = tree.get_from_path(
        "t1.nex",
        "nexus",
        taxon_namespace=tns)
tree2 = tree.get_from_path(
        "t2.nex",
        "nexus",
        taxon_namespace=tns)
tree1.encode_bipartitions()
tree2.encode_bipartitions()
print(treecompare.weighted_robinson_foulds_distance(tree1, tree2))